ANCOVA without between-subject variable

I wonder whether it's meaningful to include a covariate if I am doing an repeated-measure analysis without any between-subject variable.

In my example, I have 2 within-subject variables (island/nonisland sentence structure & plausible/implausible semantics), and the dependent variable is reading time (as the attachment). (Please ignore the very small sample size)

I don't think adding subjects' proficiency (variable: QPT) makes sense in this case because what I am comparing is the different conditions experienced by the same participant(that is, across columns, not across rows).

Am I right?